Open jeremiedb opened 7 months ago
I agree that uppercasing is a bit surprising, maybe we re-think/revise this or offer a kwarg to replace label_clean by a customer function, e.g. x -> string(x)
or identify
.
Meanwhile there's an easy solution to this that you can implement yourself, assuming that you are using DataFrames
mycolumns(df) = [Column(string(name), sortable = true, label = string(name)) for name in names(df)]
puredatatable(df) = DataTable(df, DataTableOptions(columns = mycolumns(df)))
dt = puredatatable(df)
Not a bug per se as it seems a deisgn choice, but the introduction of the override of column names in
DataTableOptions
in v0.22.9: https://github.com/GenieFramework/StippleUI.jl/blob/584f88f756b4b35ec292811eac3db613f1b10362/src/Tables.jl#L158 happened to be undesirable for our applications. I would personally favor in general limited stylistic side effects from functions by default. Could adding a kw arg inDataTableOptions
for specifying custom column names asVector{String}
cover those custom names needs, or having an opt-in "clean_colnames" boolean indicator?